Parkinson's disease (PD) is a neurodegenerative disorder that leads to deficits in both visual perception and locomotion. Research to date has neglected to determine whether PD affects the perception of a unique and important visual cue, that of human body motion (biological motion). PD patients' ability to perceive biological motion may be tied to their locomotion deficits, a link suggested by a common coding system between motor function and action perception that is mediated by frontal, motor, and parietal cortices known to be dysfunctional in PD. This perception-action link also suggests that visual training using human motion may be effective in rehabilitating the locomotion deficits in PD. The goals of the proposed study are to investigate biological motion perception and its relation to locomotion in PD (Aim1), and to determine whether visual training using biological motion can improve locomotion deficits in PD (Aim 2). Aim 1 will measure PD patients' sensitivity to perceiving biological motion characterized by different walking speeds and locomotion patterns. The hypotheses are that PD patients will be impaired at perceiving healthy (faster) walking speeds and locomotion patterns, and that patients' locomotion deficits will significantly predict the extet of their impairment in biological motion perception. Aim 2 will use a visual training paradigm that will help patients to discriminate between abnormal, inefficient locomotion and healthy, efficient locomotion. Patients' locomotion will be measured naturalistically at home using activity monitoring in order to determine whether the visual training leads to functional changes in locomotion in patients' daily lives. Together, these goals will enable researchers and clinicians to better understand how PD affects the visual analysis of other humans and their actions, a perceptual ability that is critical to effective social functioning. It will also allow researchersto elucidate the perception-action link in PD, including the specific relation between real-world locomotion and biological motion perception, and how this is affected by PD-related neural dysfunction. Finally, it will provide data regarding the effectiveness of a novel intervention that may improve patients' disordered locomotion. These data will be critically important in establishing whether visual training can be used as a rehabilitation technique in PD that would improve patients' natural walking ability in daily life.